Replication Material 'Cutting trees to balance budgets: IMF programs are associated with increased deforestation'

Download all the files and save locally. Open the R Notebook 'IMFFOR_Dec2025_Replication.Rmd' for full replication of datasets, figures, and analyses. 

For the text analysis, the following files are relevant:
* IMFMonitor2023_Conditions_Raw.csv -- IMF Conditionality raw text of policy reforms
* lemmatization-en.txt -- Lemmatization
* lemma_IMFConditions_Dec25.rds -- Lemma/Tokens object
* dfm_IMFConditions_Dec25.rds -- DFM
* dict_output_Dec25.rds -- IMF forest conditionality, as derived using a dictionary approach

For the regression analyses and figures, use the following files: 
* IMFFOR_sample.tab -- Sample of countries (low- and middle-income countries with forest cover of at least 1% in 2000)
* treecover_loss_from_fires__ha_GFW_20251214.tab -- Dependent variable: Tree cover loss, net of fires
* IMF55_1980-2019.tab -- Independent variable: IMF program dummy
* IMFMonitor2023_Conditions_Main.tab -- Independent variable: IMF condition counts
* dt_WDI_Dec2025.rds -- Controls: WDI
* dataset_T20_23_57.803850913Z_IMF.RES_WEO_9.0.0.tab -- Controls: WEO
* DreherSturmVreeland2009_UNSCMembership.tab -- Controls: UN Security Council membership
* UNGAvoting_v28_20210916.rds -- Controls: UNGA voting affinity
* Wuepper2024_POLICY DATABASE 1.0.dta -- Controls: Forest policies

Alternatively, load the fully compiled dataset used for the regression analysis:
* dt_IMFFOR_Dec2025_final.rds